Clustering association rules

Author

Lent, B. ; Swami, Arun ; Widom, Jennifer

Author_Institution

Dept. of Comput. Sci., Stanford Univ., CA, USA

fYear

1997

fDate

7-11 Apr 1997

Firstpage

220

Lastpage

231

Abstract

The authors consider the problem of clustering two-dimensional association rules in large databases. They present a geometric-based algorithm, BitOp, for performing the clustering, embedded within an association rule clustering system, ARCS. Association rule clustering is useful when the user desires to segment the data. They measure the quality of the segmentation generated by ARCS using the minimum description length (MDL) principle of encoding the clusters on several databases including noise and errors. Scale-up experiments show that ARCS, using the BitOp algorithm, scales linearly with the amount of data

Keywords

data analysis; errors; noise; pattern recognition; transaction processing; very large databases; 2D association rule clustering; ARCS; BitOp geometric-based algorithm; data segmentation; encoding; errors; large databases; minimum description length principle; noise; scale-up experiments; segmentation quality; Association rules; Clustering algorithms; Computer science; Dairy products; Data mining; Demography; Length measurement; Spatial databases; Transaction databases; Visual databases;

fLanguage

English

Publisher

ieee

Conference_Titel

Data Engineering, 1997. Proceedings. 13th International Conference on

Conference_Location

Birmingham

ISSN

1063-6382

Print_ISBN

0-8186-7807-0

Type

conf

DOI

10.1109/ICDE.1997.581756

Filename

581756